CN114677574B - Method and system for diagnosing image fault for automatic driving - Google Patents
Method and system for diagnosing image fault for automatic driving Download PDFInfo
- Publication number
- CN114677574B CN114677574B CN202210578248.0A CN202210578248A CN114677574B CN 114677574 B CN114677574 B CN 114677574B CN 202210578248 A CN202210578248 A CN 202210578248A CN 114677574 B CN114677574 B CN 114677574B
- Authority
- CN
- China
- Prior art keywords
- image
- pixel points
- abnormal pixel
- preset threshold
- variance
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Analysis (AREA)
Abstract
The present disclosure relates to a method of diagnosing image failure for automatic driving, including: sliding the diagnosis window to scan the image to be diagnosed; and after sliding the diagnosis window every time, judging whether an image fault exists according to the number of the abnormal pixel points in the current diagnosis window and the average value and the variance of Euclidean distances among all the abnormal pixel points. The image fault diagnosis scheme is particularly suitable for deep learning of an automatic driving system, and innovatively provides evaluation criteria and a method for image faults of automatic driving, so that the calculated amount is reduced, the diagnosis reliability is improved, and the automatic driving cruise performance of a vehicle is improved.
Description
Technical Field
The present disclosure relates to the field of computer vision technology, and more particularly, to a method and system for image fault diagnosis, for example, for automated driving.
Background
Cameras are an important component of autonomous driving systems as a visual component of autonomous vehicles. The automatic driving system acquires various feature quantities from the camera image as a cruising basis for automatic driving. In the image processing process of the automatic driving system, whether the image has a fault or not needs to be diagnosed at first, and reliable basis and guarantee are provided for subsequent processing.
In the prior art, image fault diagnosis focuses more on detecting defective pixels, aiming at identifying defective pixels and the types of the defective pixels. However, the current defective pixel diagnosis method is not satisfactory in practical application, and there is a need and a space for further improvement because false alarm occurs often due to a large amount of calculation and a complex diagnosis system, which affects the reliability of diagnosis.
Disclosure of Invention
The present disclosure provides a method of diagnosing image failure for automatic driving, including:
sliding the diagnosis window to scan the image to be diagnosed;
and after the diagnosis window is slid every time, judging whether an image fault exists or not according to the number of the abnormal pixel points in the current diagnosis window and the average value and the variance of Euclidean distances among all the abnormal pixel points.
Advantageously, after each sliding of the diagnostic window, the following steps are performed:
counting the number of abnormal pixel points in the current diagnosis window;
if the number of the abnormal pixel points is less than or equal to a preset threshold value of the number of the abnormal pixel points, continuously sliding the diagnosis window to scan the rest part of the image;
if the number of the abnormal pixel points is larger than a preset threshold of the number of the abnormal pixel points, calculating the average value and the variance of Euclidean distances among all the abnormal pixel points, generating image fault information when the average value is smaller than the preset threshold of the average value and the variance is smaller than a preset threshold of the variance, and continuously sliding the diagnosis window to scan the rest part of the image when the average value is larger than or equal to the preset threshold of the average value or the variance is larger than or equal to the preset threshold of the variance.
Advantageously, after each sliding of the diagnostic window, the following steps are performed:
counting the number of abnormal pixel points in the current diagnosis window, and calculating the average value and the variance of Euclidean distances among all the abnormal pixel points;
if the number of the abnormal pixel points is larger than a preset threshold of the number of the abnormal pixel points, the average value is smaller than a preset threshold of the average value, and the variance is smaller than a preset threshold of the variance, generating image fault information;
if the number of the abnormal pixel points is smaller than or equal to a preset threshold of the number of the abnormal pixel points, or the average value is larger than or equal to a preset threshold of the average value, or the variance is larger than or equal to a preset threshold of the variance, the diagnosis window is continuously slid to scan the rest part of the image.
Advantageously, the length and width of the diagnostic window are set in accordance with a deep learning minimum identification region.
The method has the advantage that when the gray value of the pixel point in the image is not within the gray threshold range of the pixel point, the pixel point is judged to be an abnormal pixel point.
The present disclosure also provides a system for diagnosing image failure for automatic driving, including:
an image acquisition module arranged to receive images acquired in real time;
an image scanning module provided with a diagnosis window and scanning the image by sliding the diagnosis window;
and the fault diagnosis module judges whether an image fault exists according to the number of the abnormal pixel points in the current diagnosis window and the average value and the variance of Euclidean distances among all the abnormal pixel points after sliding the diagnosis window every time.
Advantageously, the fault diagnosis module is configured to perform the following diagnostic steps after each sliding of the diagnostic window:
counting the number of abnormal pixel points in the current diagnosis window;
if the number of the abnormal pixel points is less than or equal to a preset threshold value of the number of the abnormal pixel points, continuously sliding the diagnosis window to scan the rest part of the image;
if the number of the abnormal pixel points is larger than a preset threshold of the number of the abnormal pixel points, calculating the average value and the variance of Euclidean distances among all the abnormal pixel points, generating image fault information when the average value is smaller than the preset threshold of the average value and the variance is smaller than a preset threshold of the variance, and continuously sliding the diagnosis window to scan the rest parts of the image when the average value is larger than or equal to the preset threshold of the average value or the variance is larger than or equal to the preset threshold of the variance.
Advantageously, the fault diagnosis module is configured to, after each sliding of the diagnosis window, perform the following steps:
counting the number of abnormal pixel points in the current diagnosis window, and calculating the average value and the variance of Euclidean distances among all the abnormal pixel points;
if the number of the abnormal pixel points is larger than a preset threshold of the number of the abnormal pixel points, the average value is smaller than a preset threshold of the average value, and the variance is smaller than a preset threshold of the variance, generating image fault information;
and if the number of the abnormal pixel points is less than or equal to a preset threshold of the number of the abnormal pixel points, or the average value is greater than or equal to a preset threshold of the average value, or the variance is greater than or equal to a preset threshold of the variance, continuing to slide the diagnosis window to scan the rest part of the image.
The present disclosure also provides an electronic device comprising a processor and instructions stored thereon for execution by the processor, wherein the instructions, when executed by the processor, implement the steps of the method for diagnosing image faults according to any of the above embodiments.
The present disclosure also provides a storage medium storing an application program, which when executed by a processor, implements the steps of the method for diagnosing image failure according to any one of the above embodiments.
The image fault diagnosis scheme is particularly suitable for deep learning of an automatic driving system. The inventor discovers that the automatic driving deep learning has higher fault tolerance on abnormal pixel points and only a specific dead pixel state can influence the accuracy of the visual deep learning based on analysis, thereby creating an evaluation standard and a method for image faults. The method can effectively reduce the calculated amount of the automatic driving system, improve the reliability of image diagnosis and further improve the automatic driving cruising performance of the vehicle.
Drawings
FIG. 1 is a flow chart of an exemplary method for diagnosing image faults.
FIG. 2 is a specific embodiment of the process shown in FIG. 1.
Fig. 3-1 through 3-3 are exemplary camera images to be diagnosed.
Fig. 4 is an exemplary system for diagnosing image faults.
Detailed Description
To further describe the relevant content of the present disclosure, the following description will be made in conjunction with a plurality of embodiments and accompanying drawings. The description relates to only some embodiments and, on the basis of the described embodiments, a person skilled in the art will be able to conceive of other undescribed related or similar embodiments of said embodiments without the need for inventive work.
FIG. 1 shows one example method for diagnosing camera image faults, which includes the steps of:
s101: the diagnostic window is slid to scan the image to be diagnosed.
In one or more embodiments, the diagnosis window is set to have a suitable length m and width n, the values of m and n being related to the accuracy of the image failure diagnosis, the smaller the value, the higher the accuracy, and the larger the calculation amount. For this reason, the minimum recognition area may be set depending on the machine deep learning and/or the size of the image to be detected. For example, the size of the diagnostic window is between the machine deep learning minimum identification region and the size of the image to be detected.
In addition, the setting of the diagnostic window may further include each sliding step of the diagnostic window, i.e., the distance moved by the diagnostic window each time the diagnostic window slides. For example, in one or more embodiments, the diagnostic window is slid only one pixel in one direction at a time to ensure that each pixel on the image is scanned.
The step of sliding the diagnostic window is mainly based on the set diagnostic window scanning the image to be diagnosed according to a predetermined sliding path, and selecting a target area for the subsequent diagnostic step. For example, the diagnostic window slides on the image in the horizontal direction according to the set step length, and after one line of scanning is completed, the diagnostic window slides in the vertical direction, that is, the image to be diagnosed is scanned line by line from left to right and from top to bottom until the whole image is traversed and no defect or fault is found.
S102: and carrying out image fault diagnosis after sliding the diagnosis window every time.
Specific contents of this step include, but are not limited to: and after the diagnosis window is slid according to a preset step length each time, judging whether an image fault exists according to the number of the abnormal pixel points in the current diagnosis window and the average value and the variance of Euclidean distances among all the abnormal pixel points. In other words, after each sliding of the diagnostic window is performed according to the preset step length, image fault diagnosis is performed, and the factors considered in the diagnostic process at least include one or more of the number of abnormal pixel points in the diagnostic window and the average value and variance of euclidean distances between all the abnormal pixel points.
In one or more embodiments, whether an abnormal pixel exists is determined according to the gray value of each pixel of the image. For example, the gray scale value range of the image may be set as the threshold range of the abnormal pixel according to the pixel characteristics of the specific camera, and when the gray scale value of the pixel is not within the threshold range, the pixel is determined to be the abnormal pixel.
The calculation method of the euclidean distance between the abnormal pixel points is as follows (formula 1, for example, accurate to three bits after decimal point):
wherein, the first and the second end of the pipe are connected with each other,
is a point
And point
The euclidean distance between them,jis as followsjThe abnormal pixel points are selected from the group of abnormal pixel points,kis as followskEach abnormal pixel point;
the average value of the euclidean distances is calculated as follows (formula 2):
the variance calculation method of the euclidean distance is as follows (formula 3):
wherein the content of the first and second substances,i= 
，Nthe number of abnormal pixels.
FIG. 2 illustrates a specific embodiment of the method shown in FIG. 1, which includes:
s1: a camera image acquired in real time is acquired.
S2: and sliding the diagnosis window by a preset step length according to a preset sliding path to scan the image. As described above, the sliding path of the diagnosis window is, for example, from left to right and from top to bottom, and the images to be diagnosed slide sequentially.
S3: and after the diagnosis window slides for a preset step length, counting the number of the abnormal pixel points in the current diagnosis window and the coordinates of each abnormal pixel point.
S4: and if the number of the abnormal pixel points is less than or equal to a preset threshold value, continuously sliding the diagnosis window to scan the rest part of the image until the whole image is scanned to finish diagnosis of a single image, and starting acquisition and diagnosis of the next image. The preset threshold of the number of abnormal pixels can be set as required, wherein the larger the preset threshold of the number of abnormal pixels is, the lower the diagnosis precision is, and the smaller the calculation amount is.
S5: if the number of the abnormal pixel points is larger than a preset threshold value of the number of the abnormal pixel points, calculating the average value and the variance of Euclidean distances among all the abnormal pixel points;
s6: if the average value and the variance are both smaller than a preset threshold value, generating image fault information, indicating image errors, completing the diagnosis of the current image, and starting the acquisition and diagnosis of the next image; and if the average value and the variance are both larger than or equal to the preset threshold value, continuing sliding the diagnosis window to scan the rest part of the image until the whole image is scanned to finish diagnosis of a single image, and starting acquisition and diagnosis of the next image. The Euclidean distance average value preset threshold and the Euclidean distance variance preset threshold can also be set according to requirements, wherein the larger the threshold is, the higher the diagnosis precision is, and the smaller the calculation amount is.
The following will specifically describe the related content of the above method by taking a plurality of images of the camera to be diagnosed shown in fig. 3-1, 3-2, and 3-3 as an example, wherein the black squares in the figure represent abnormal pixel points.
For the diagnosis method, for example, the width and length of the diagnosis window are set to 25 × 25 pixels, and a preset threshold a =3 of the number of abnormal pixel points is set, a preset threshold b =4 of the euclidean distance average of the abnormal pixel points, and a preset threshold c =3 of the euclidean distance variance.
Firstly, an image (as shown in fig. 3-1) is collected, the number of abnormal pixel points in the image and the coordinates thereof are counted, wherein the number of the abnormal pixel points is 2, and the coordinates are (5, 25) and (8, 19), respectively. Since the number of the abnormal pixel points in fig. 3-1 is less than the preset threshold value a =3, and the whole image is scanned, the image is determined to be free of defects, and the diagnosis of the current image is completed.
Then, the image is collected again (as shown in fig. 3-2), and the number of abnormal pixel points and the coordinates thereof in the image are counted, wherein the number of abnormal pixel points is 4, and the coordinates are (6, 23), (20, 23), (8, 15), (21, 15), respectively. Since the number of the abnormal pixel points in fig. 3-2 is greater than the preset threshold value a =3, the euclidean distances between the abnormal pixel points are calculated according to the formula (1), and the obtained euclidean distances are respectively: 3.742, 8.246, 18.788, 14.422, 8.062, 13. Then, the mean euclidean distance is calculated to be 11.043 according to the formula (2), and the variance euclidean distance is calculated to be 24.208 according to the formula (3). Since the average value of fig. 3-2 is greater than the preset value b =4, the variance is greater than the preset value c =3, and the entire picture is scanned, it is assumed that the image is defect-free.
And then, collecting the image again (as shown in fig. 3-3), and counting the number of abnormal pixel points and the coordinates thereof in the image, wherein the number of the abnormal pixel points is 4, and the coordinates are (18, 13), (15, 13), (14, 17) and (15, 17), respectively. Since the number of the abnormal pixel points in fig. 3-3 is greater than the preset threshold value a =3, the euclidean distances between all the abnormal pixel points are calculated according to the formula (1), and the obtained euclidean distances are respectively: 3. 5.657, 5, 4.123, 4, 1. The mean value of euclidean distances calculated according to equation (2) was 3.797, and the variance of euclidean distances calculated according to equation (3) was 2.252. The mean of fig. 3-3 is less than the preset value b =4 and the variance of fig. 3-3 is less than the preset value c =3, indicating an image error, generating a malfunction alarm signal, reacquiring the image, and entering the next diagnostic cycle.
The specific embodiments described above may have a variety of variations, including, for example:
s1: camera images are acquired in real time.
S2: the diagnostic window is slid by a predetermined step size to scan the acquired image according to a predetermined sliding path. As described above, the sliding path of the diagnosis window is, for example, from left to right and from top to bottom, and the images to be diagnosed slide sequentially.
S3: and after the diagnosis window is slid by a preset step length, counting the number of abnormal pixel points in the current diagnosis window and the coordinates of each abnormal pixel point, and calculating the average value and the variance of Euclidean distances among all the abnormal pixel points.
S4: if the number of the abnormal pixel points is larger than a preset threshold value and the average value and the variance are both smaller than the preset threshold value, generating image fault information, completing the diagnosis of the current image, and starting the acquisition and diagnosis of the next image;
s5: if the number of the abnormal pixel points is smaller than or equal to a preset threshold, or the average value is larger than or equal to a preset threshold of the average value, or the variance is larger than or equal to a preset threshold of the variance, the diagnosis window is continuously slid to scan the rest parts of the image until the whole image is scanned to finish diagnosis of a single image, and acquisition and diagnosis of the next image are started.
Fig. 4 illustrates a system 10 for image fault diagnosis, comprising:
an image acquisition module 11 arranged to receive camera images acquired in real time;
an image scanning module 12 which is provided with a diagnosis window and scans the camera image by sliding the diagnosis window;
and the fault diagnosis module 13 is used for judging whether an image fault exists or not according to the number of the abnormal pixel points in the current diagnosis window and the average value and the variance of Euclidean distances among all the abnormal pixel points after sliding the diagnosis window every time.
For example, in one or more embodiments, the fault diagnosis module 13 is configured to:
after the diagnosis window is slid every time, counting the number of abnormal pixel points in the current diagnosis window;
if the number of the abnormal pixel points is less than or equal to a preset threshold value of the number of the abnormal pixel points, continuously sliding the diagnosis window to scan the rest part of the image;
if the number of the abnormal pixel points is larger than a preset threshold of the number of the abnormal pixel points, calculating the average value and the variance of Euclidean distances among all the abnormal pixel points, generating image fault information when the average value is smaller than the preset threshold of the average value and the variance is smaller than a preset threshold of the variance, and continuously sliding the diagnosis window to scan the rest part of the image when the average value is larger than or equal to the preset threshold of the average value or the variance is larger than or equal to the preset threshold of the variance.
For another example, in one or more embodiments, the fault diagnosis module 13 is configured to, after sliding the diagnosis window each time, count the number of abnormal pixel points in the current diagnosis window, and calculate an average value and a variance of euclidean distances between all the abnormal pixel points;
if the number of the abnormal pixel points is larger than a preset threshold of the number of the abnormal pixel points, the average value is smaller than a preset threshold of the average value, and the variance is smaller than a preset threshold of the variance, generating image fault information;
if the number of the abnormal pixel points is smaller than or equal to a preset threshold of the number of the abnormal pixel points, or the average value is larger than or equal to a preset threshold of the average value, or the variance is larger than or equal to a preset threshold of the variance, the diagnosis window is continuously slid to scan the rest part of the image.
It will be understood by those skilled in the art that all or part of the steps in the methods of the embodiments described above may be implemented by hardware related to instructions of a computer program, and the computer program may be stored in a non-volatile computer readable storage medium, and when executed, the computer program may include the procedures of the embodiments of the methods described above, which are not described in detail herein.
The above examples are only for illustrating the technical solutions of the present disclosure, and not for limiting the same. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. A method of diagnosing image faults for autonomous driving, comprising:
sliding the diagnosis window to scan the image to be diagnosed;
after the diagnosis window is slid every time, counting the number of abnormal pixel points in the current diagnosis window;
if the number of the abnormal pixel points is less than or equal to the preset threshold of the number of the abnormal pixel points, continuously sliding the diagnosis window to scan the rest part of the image;
if the number of the abnormal pixel points is larger than a preset threshold of the number of the abnormal pixel points, calculating the average value and the variance of Euclidean distances among all the abnormal pixel points, generating image fault information when the average value is smaller than the preset threshold of the average value and the variance is smaller than a preset threshold of the variance, and continuously sliding the diagnosis window to scan the rest part of the image when the average value is larger than or equal to the preset threshold of the average value or the variance is larger than or equal to the preset threshold of the variance.
2. The method of claim 1, wherein the length and width of the diagnostic window are set according to a deep learning minimum identification region.
3. The method of claim 1, wherein when the gray level value of a pixel in the image is not within the gray level threshold range of the pixel, the pixel is determined to be an abnormal pixel.
4. A method of diagnosing image faults for autonomous driving, comprising:
sliding the diagnosis window to scan the image to be diagnosed;
after the diagnosis window is slid every time, counting the number of abnormal pixel points in the current diagnosis window, and calculating the average value and variance of Euclidean distances among all the abnormal pixel points;
if the number of the abnormal pixel points is larger than a preset threshold of the number of the abnormal pixel points, the average value is smaller than a preset threshold of the average value, and the variance is smaller than a preset threshold of the variance, generating image fault information;
and if the number of the abnormal pixel points is less than or equal to a preset threshold of the number of the abnormal pixel points, or the average value is greater than or equal to a preset threshold of the average value, or the variance is greater than or equal to a preset threshold of the variance, continuing to slide the diagnosis window to scan the rest part of the image.
5. The method of claim 4, wherein the length and width of the diagnostic window are set according to a deep learning minimum identification region.
6. The method of claim 4, wherein when the gray level value of a pixel in the image is not within the gray level threshold range of the pixel, the pixel is determined to be an abnormal pixel.
7. A system for diagnostic image faults for autonomous driving, comprising:
an image acquisition module arranged to receive images acquired in real time;
an image scanning module provided with a diagnosis window and scanning the image by sliding the diagnosis window;
a fault diagnosis module configured to perform the following diagnostic steps after each sliding of the diagnostic window:
counting the number of abnormal pixel points in the current diagnosis window;
if the number of the abnormal pixel points is less than or equal to the preset threshold of the number of the abnormal pixel points, continuously sliding the diagnosis window to scan the rest part of the image;
if the number of the abnormal pixel points is larger than a preset threshold of the number of the abnormal pixel points, calculating the average value and the variance of Euclidean distances among all the abnormal pixel points, generating image fault information when the average value is smaller than the preset threshold of the average value and the variance is smaller than a preset threshold of the variance, and continuously sliding the diagnosis window to scan the rest part of the image when the average value is larger than or equal to the preset threshold of the average value or the variance is larger than or equal to the preset threshold of the variance.
8. A system for diagnostic image faults for autonomous driving, comprising:
an image acquisition module arranged to receive images acquired in real time;
an image scanning module provided with a diagnosis window and scanning the image by sliding the diagnosis window;
a fault diagnosis module configured to, after each sliding of the diagnosis window, perform the steps of:
counting the number of abnormal pixel points in the current diagnosis window, and calculating the average value and the variance of Euclidean distances among all the abnormal pixel points;
if the number of the abnormal pixel points is larger than a preset threshold of the number of the abnormal pixel points, the average value is smaller than a preset threshold of the average value, and the variance is smaller than a preset threshold of the variance, generating image fault information;
and if the number of the abnormal pixel points is less than or equal to a preset threshold of the number of the abnormal pixel points, or the average value is greater than or equal to a preset threshold of the average value, or the variance is greater than or equal to a preset threshold of the variance, continuing to slide the diagnosis window to scan the rest part of the image.
9. An electronic device, comprising:
a processor;
a memory storing instructions for execution by the processor;
wherein the instructions, when executed by the processor, implement the method of any of claims 1 to 6.
10. A storage medium, characterized in that it stores an application program which, when executed by a processor, implements the method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210578248.0A CN114677574B (en) | 2022-05-26 | 2022-05-26 | Method and system for diagnosing image fault for automatic driving |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210578248.0A CN114677574B (en) | 2022-05-26 | 2022-05-26 | Method and system for diagnosing image fault for automatic driving |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114677574A CN114677574A (en) | 2022-06-28 |
| CN114677574B true CN114677574B (en) | 2022-10-21 |
Family
ID=82079966
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210578248.0A Active CN114677574B (en) | 2022-05-26 | 2022-05-26 | Method and system for diagnosing image fault for automatic driving |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114677574B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107172421A (en) * | 2017-07-17 | 2017-09-15 | 深圳森阳环保材料科技有限公司 | A kind of picture quality diagnostic system for video monitoring |
| CN107451975A (en) * | 2017-04-25 | 2017-12-08 | 中国人民解放军空军工程大学 | A kind of view-based access control model weights similar picture quality clarification method |
| CN109167997A (en) * | 2017-03-30 | 2019-01-08 | 安徽四创电子股份有限公司 | A kind of video quality diagnosis system and method |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104408705B (en) * | 2014-09-23 | 2017-10-20 | 西安电子科技大学 | A kind of method for detecting abnormality of high spectrum image |
| CN107016669B (en) * | 2017-03-27 | 2019-08-23 | 福州瑞芯微电子股份有限公司 | A kind of dead pixel points of images detection method and device |
| WO2020010638A1 (en) * | 2018-07-13 | 2020-01-16 | 华为技术有限公司 | Method and device for detecting defective pixel in image |
| CN112365435B (en) * | 2019-07-25 | 2023-02-10 | 天津大学青岛海洋技术研究院 | Image dead pixel detection and edge-based restoration algorithm |
| CN110650334B (en) * | 2019-10-29 | 2021-01-22 | 锐芯微电子股份有限公司 | Dead pixel detection and correction method and device, storage medium and terminal |
| WO2021207879A1 (en) * | 2020-04-13 | 2021-10-21 | 深圳市大疆创新科技有限公司 | Image processing method and apparatus, electronic device and storage medium |
| CN113674238B (en) * | 2021-08-16 | 2024-06-11 | 浙江大华技术股份有限公司 | Method and device for detecting dead pixel, electronic equipment and storage medium |
| CN114202539A (en) * | 2021-12-18 | 2022-03-18 | 南京理工大学 | Hyperspectral image anomaly detection method based on end-to-end RX |
-
2022
- 2022-05-26 CN CN202210578248.0A patent/CN114677574B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109167997A (en) * | 2017-03-30 | 2019-01-08 | 安徽四创电子股份有限公司 | A kind of video quality diagnosis system and method |
| CN107451975A (en) * | 2017-04-25 | 2017-12-08 | 中国人民解放军空军工程大学 | A kind of view-based access control model weights similar picture quality clarification method |
| CN107172421A (en) * | 2017-07-17 | 2017-09-15 | 深圳森阳环保材料科技有限公司 | A kind of picture quality diagnostic system for video monitoring |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114677574A (en) | 2022-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP4816725B2 (en) | Image processing apparatus, image processing program, electronic camera, and image processing method for image analysis of lateral chromatic aberration | |
| CN101271333B (en) | Localization method for a moving robot | |
| CN112598922B (en) | Parking space detection method, device, equipment and storage medium | |
| CN103759758A (en) | Method for detecting position of automobile meter needle based on mechanical angle and scale identification | |
| US11928805B2 (en) | Information processing apparatus, information processing method, and storage medium for defect inspection and detection | |
| CN108492327B (en) | AOI image matching method, system, readable storage medium and intelligent device | |
| CN110197180B (en) | Character defect detection method, device and equipment | |
| CN112115913B (en) | Image processing method, device and equipment and storage medium | |
| CN111383246B (en) | Scroll detection method, device and equipment | |
| CN113256729A (en) | External parameter calibration method, device, equipment and storage medium for laser radar and camera | |
| CN114677574B (en) | Method and system for diagnosing image fault for automatic driving | |
| CN110706257B (en) | Identification method of effective characteristic point pair, and camera state determination method and device | |
| EP4375924A1 (en) | Cell alignment degree measurement method, controller, detection system, and storage medium | |
| CN114529543B (en) | Installation detection method and device for peripheral screw gasket of aero-engine | |
| CN112214629A (en) | Loop detection method based on image recognition and movable equipment | |
| CN111582332A (en) | Picture identification method for dropper component of high-speed railway contact network | |
| CN116128853A (en) | Production line assembly detection method, system, computer and readable storage medium | |
| CN112292847A (en) | Image processing apparatus, mobile apparatus, method, and program | |
| CN115931895A (en) | Method and device for checking defects of power transmission line | |
| CN113326749B (en) | Target detection method and device, storage medium and electronic equipment | |
| CN115861161A (en) | Machine learning system, learning data collection method, and storage medium | |
| CN114022873A (en) | Instrument state detection method, electronic device and storage medium | |
| KR20230024160A (en) | System and method for semantic segmentation learning data error detection and correction | |
| CN113870754A (en) | Method and system for judging defects of panel detection electronic signals | |
| CN113361413A (en) | Mileage display area detection method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |